1 Introduction

Renewable energy sources, such as solar, wind and geothermal, have been developed in large scale to mitigate the increasingly serious oil shortage and air pollution [1,2,3]. The large-scale energy-storage technology is one of the critical factors for the development of renewable energy resources. Among them, lithium-ion battery (LIB) as an energy-storage device has drawn discernible attention that can meet the applications with its low self-discharge, high energy density and long cycle life [4,5,6]. Graphite is the most widely used commercial anode material for LIB due to its low cost, good specific capacity and cyclic stability [7,8,9]. Nevertheless, due to the working voltage of graphite under 1.0 V (vs. Li+/Li), the nonconductive solid–electrolyte interphase film would be generated on the material surface. Furthermore, the graphite material also along with huge volume changes (~ 9%) from C to LixC6 (0 < x < 6) process. The issues would cause a capacity loss and potential safety issue in actual use.

To solve the above problems, plenty of anode materials of LIBs have been investigated. In particular, the transition metal oxides can acquire and deliver high reversible capacities and work in high potential for a high-energy LIBs, which have been investigated as electrode materials [10,11,12,13,14,15]. Among them, orthorhombic niobium oxide is thought to be a promising anode material for LIBs based on the considerable theoretical capacity with multiple redox couples, minimal volume change (3%) upon cycling and excellent rate capability [11, 16, 17]. Also, the appropriate voltage of niobium oxide (1.0–2.0 V vs. Li+/Li) can effectively prevent the formation of Li dendrites and improve the safety of LIBs. However, the moderate Li+ diffusion dynamics and poor electrical conductivity (3.4 × 10−6 S·cm−1 at 300 K) of bulk Nb2O5 material greatly limit its high rate capability, which is a critical factor for the application of LIBs. Therefore, great efforts have been explored to promote the ion diffusion rate and elevate the electrical conductivity of Nb2O5 electrodes. Well-designed nanostructures of electrodes can effectively shorten the distance of Li+ diffusion. For instance, Lou et al. [18] found that three-dimensionally ordered macroporous orthorhombic Nb2O5 prepared by hard template method exhibited a remarkably high rate performance. Fu et al. [19] fabricated yolk–shell-structured Nb2O5 microspheres via spray-drying method to enhance the electronic conductivity and rate capability of the electrode. However, the high cost, complex synthetic process and harsh preparative conditions of Nb2O5 are restricting the practical perspective.

Apart from Li+ diffusion, the poor electric conductivity of Nb2O5 electrode is the other important constraint for rate performance. Coating conductive materials on the electrode is an efficient way to facilitate charge migration. It is worth noting that the heteroatom (N, B, S, and P)-doped carbon material with high electrical conductivity is an ideal candidate for coated layer [20,21,22]. Nitrogen element has been widely introduced into the carbon material; it can not only create vast defects and active sites to motivate the redox operation, but also enhance the electronic conductivity with the faster reaction kinetics. Furthermore, an ionic liquid was extensively used as raw materials for nitrogen and carbon to obtain N-doped graphitized carbon. Owing to the fluidic properties of ionic liquid, it can more easily penetrate porous materials than the conventional carbon sources and then can form the thin uniform coating layer on the materials surface, which would not change the porous surface characteristics [23, 24]. Therefore, ionic liquid-induced N-doped graphitized carbon coating has shown a prominent prospect for solving the poor electric conductivity of Nb2O5 electrode.

Herein, we report the ultrathin N-doped carbon-coated porous T-Nb2O5 microsphere prepared through a simple and convenient thermal treatment strategy. N and C are introduced to the electrode material through the impregnation of ionic liquid. N-doped carbon networks can offer high electrical conductivity. On the other hand, the porous structure can shorten the transfer path of electrons and ions to further promote the rate performance. Therefore, the as-synthesized T-Nb2O5@N–C demonstrates a high specific capacity (221.7 mAh·g−1 at 1C) and superior rate capability (155.5 mAh·g−1 at 20C). The incorporation of an ionic liquid as the coating layer material and structural modification of T-Nb2O5 to improve the excellent long cycling life is an imperative strategy, which is suitable for being widely used in electrode material design (Scheme 1).

Scheme 1
scheme 1

Synthesis processes of N-doped carbon wrapped T-Nb2O5 microsphere

Full size image

2 Experimental

2.1 Preparation of T-Nb2O5

About 2.7 g NbCl5 was completely dissolved to 80 ml ethylene glycol with vigorous stirring. The solution was then retained at 200 °C for 24 h by the solvothermal method. Subsequently, the precursor precipitate was collected and washed with water and ethanol several times. After that, the precursor was put in a muffle furnace and heated at 600 °C for 2 h to obtain the final products T-Nb2O5.

2.2 Preparation of T-Nb2O5@N–C

At room temperature, 100 μl ionic liquid (1-ethyl-3-methylimidazolium dicyanamide), 0.4 g T-Nb2O5 and 10 ml acetone were mixed with vigorous stirring until the evaporation of the solvent. After that, the mixture was calcined under N2 at 600 °C for 2 h to obtain the T-Nb2O5@N–C.

2.3 Characterizations

X-ray diffractometer (XRD, Dmax/2500PC) was used to determine the crystalline phase of the materials with a scanning rate of 2 (°)·min−1. The morphologies and microstructures were surveyed using a scanning electron microscope (SEM, ZEISS SIGMA 500) and a field emission transmission electronic microscope (FETEM, JEM-2100). Energy-dispersive X-ray spectroscopy (EDS) was employed to analyze elemental composition and distribution on the surface. The Brunauer–Emmett–Teller (BET) surface area and pore volume were identified by N2 adsorption–desorption isotherm (Micromeritics ASAP 2020 analyzer). The carbon contents in the composites were measured by thermal gravimetric analysis (TG, SDT Q600). The valence states of the key elements in samples were studied by X-ray photoelectron spectroscopy (XPS, Thermo ESCALAB 250XI electron spectrometer, Al Kα radiation). Raman spectroscopy (LabRAM HR Evolution, excited by 532-nm laser) was employed to investigate the structural characteristic of carbonaceous materials in the products. The electrical conductivity of the T-Nb2O5@N–C and T-Nb2O5 was directly measured by semiconductor powder resistivity tester (ST2722) and ultra-high-resistance microcurrent tester (ST2643).

2.4 Electrochemical measurements

The electrode sheets were composed of T-Nb2O5 or T-Nb2O5@N–C powder, acetylene black and polyvinylidene fluoride (mass ratio of 70:20:10) in the N-methyl-2-pyrrolidone. The mixtures were uniformly cast on Cu foil and were put in a vacuum oven at 120 °C overnight. The loading amount of the electrode was around 2 mg·cm−2. The full battery was constructed of commercial LiNi0.5Mn0.3Co0.2O4 and T-Nb2O5@N–C material as cathode and anode with a mass ratio of 1.4:1. The preparation process of LiNi0.5Mn0.3Co0.2O4 electrode is consistent with that of T-Nb2O5@N–C. All samples were evaluated using CR2025 coin cells with lithium metal as the counter and 1 mol·L−1 LiPF6 dissolved in ethylene carbonate (EC)/diethyl carbonate (DEC) /dimethyl carbonate(DC) (1:1:1 in volume ratio) as the electrolyte. The LANDCT2001A system was used to test the electrochemical performance with the potential range of 1.0–3.0 V at ambient temperature. The electrochemical capacities of the samples were tested with the quality of active material. Cycling voltammetry (CV) and electrochemical impedance spectroscopy (EIS) measurements were performed using a CHI 650D electrochemical workstation.

3 Results and discussion

In Fig. 1a, the crystalline phase of T-Nb2O5 and T-Nb2O5@N–C was analyzed by XRD. All the strong diffraction peaks in both samples can be indexed to orthorhombic Nb2O5 with a Pbam space group (PDF No. 27-1003) [25]. Besides, no carbon peaks can be observed due to that carbon is amorphous or the content of carbon is low. This result suggests that high-purity Nb2O5 composites for the two samples have been synthesized. To verify the presence of Nb2O5 and carbon in samples, Raman spectra are shown in Fig. 1b. The peaks in the range of 200–500 cm−1 and 695 cm−1 for the T-Nb2O5 in Fig. 1b are assigned to bending vibration and stretching vibration of the Nb–O–Nb bond, respectively [26, 27]. Meanwhile, T-Nb2O5@N–C along with the introduction of ionic liquids has two strong peaks at around 1362 and 1609 cm−1, deriving from D-band and G-band of carbon, respectively [28]. However, the characteristic peaks of the Nb–O–Nb bond have disappeared for T-Nb2O5@N–C, suggesting that the tight carbon-coating layer may screen the signal [29]. The ratio of the D-band to G-band intensities (ID/IG) is 0.92, certifying the graphitic N-doped carbon in T-Nb2O5@N–C sample [30]. The carbon content in T-Nb2O5@N–C is verified as 2.66% by TG analysis (Fig. 1c).

Fig. 1
figure 1

a XRD patterns and b Raman spectra of T-Nb2O5 and T-Nb2O5@N–C; c TG analysis of T-Nb2O5@N–C; SEM images of d, e T-Nb2O5 and f, g T-Nb2O5@N–C

Full size image

It is reported that the microstructure of the materials is closely related to the electrochemical performance. We thus investigated the morphology of T-Nb2O5 and T-Nb2O5@N–C. As displayed in Fig. 1d, e, T-Nb2O5 microsphere with the size of ~ 2.5 μm is composed of uniform primary nanoparticles (~ 50 nm). After the introduction of N-doped carbon, the size and shape of T-Nb2O5@N–C do not change obviously and the porous microsphere structure is maintained, as shown in Fig. 1f. The nitrogen adsorption–desorption isotherm (Fig. 2a) indicates BET surface area of T-Nb2O5@N–C of 59 m2·g−1, and T-Nb2O5@N–C has rich mesopores with a size of ~ 20 nm (Fig. 2b), in accord with SEM results (Fig. 1g). These results demonstrate that the porous T-Nb2O5@N–C microsphere has been synthesized. This mesoporous structure would expedite electrolyte penetration and then promote Li+ diffusion dynamics.

Fig. 2
figure 2

a Nitrogen adsorption–desorption isotherms and b pore size distribution of T-Nb2O5@N–C; TEM and HRTEM images of c, e T-Nb2O5 and d, f T-Nb2O5@N–C; g elemental mappings of T-Nb2O5@N–C

Full size image

The detailed morphology of T-Nb2O5 and T-Nb2O5@N–C was then investigated by TEM and HRTEM, as displayed in Fig. 2c–f. T-Nb2O5@N–C exhibits similar microsphere morphology with T-Nb2O5. Besides, HRTEM images in Fig. 2e, f show the interplanar distance of 0.39 nm in both of the two samples, which can be well indexed to the (001) plane of orthorhombic Nb2O5, corresponding to XRD results in Fig. 1a. The empty octahedral sites between its (001) planes can result in high rate capability [31]. In HRTEM image of T-Nb2O5@N–C (Fig. 2f), the ultrathin uniform N-doped carbon layer (~ 2 nm) is coated on Nb2O5 nanoparticles, as marked by the bright yellow line. The intimate contact of the carbon layer and Nb2O5 will facilitate charge transfer to deliver high capacity during the charge/discharge process [32]. The elemental mapping was implemented to investigate the material surface, and the elemental distribution of Nb, O, N and C is shown in Fig. 2g. EDS result of the T-Nb2O5@N–C powder is recorded in Fig. S1. N content in T-Nb2O5@N–C is 0.21%. The above results confirm the porosity and uniform distribution of N-doped carbon. As suggested from previous literature, the ultrathin and uniform coating of N-doped carbon is an ideal way for accelerating electrons transport.

XPS measurements were carried out to elucidate the surface chemical states and confirm the bonding relationship. Figure 3a, b shows the high-resolution Nb and O spectra of T-Nb2O5 and T-Nb2O5@N–C, respectively. In Fig. 3a, two predominant peaks at 207.2 and 209.9 eV are assigned to Nb 3d5/2 and Nb 3d3/2, which are Nb5+ state in both samples [33]. Compared to the O 1s spectra for T-Nb2O5, the peaks of T-Nb2O5@N–C shift to the high binding energy. Additionally, the newly formed C–O (531.6 eV) for T-Nb2O5@N–C is demonstrated in Fig. 3b, indicating the formation of C–O–Nb bonds during the thermal treatment process [34]. The C 1s spectra consist of four peaks including binding energies of 284.5, 285, 287 and 290 eV, which respectively represent the carbon functional group with C–C, C–N, C=N and C–O in Fig. 3c [35]. As displayed in Fig. 3d, N 1s spectrum can be divorced into three peaks with binding energies of 398.6, 400.8 and 403.5 eV, corresponding to pyridinic N, pyrrolic N and graphitic N, respectively [36]. The results further indicate that the N-doped carbon has been successfully coated on Nb2O5 material with ionic liquid as nitrogen and carbon source. N for electrode material was adopted to provide supererogatory reactive sites and degrade the energy barrier for Li+, which can dramatically improve the Li-ion storage properties.

Fig. 3
figure 3

a Nb 3d and b O 1 s high-resolution XPS spectra of T-Nb2O5 and T-Nb2O5@N–C; c C 1 s and d N 1 s XPS spectra of T-Nb2O5@N–C

Full size image

Figure 4a, b depicts representative galvanostatic charge/discharge profiles of T-Nb2O5 and T-Nb2O5@N–C at different current densities. The potential difference between charge and discharge curves of T-Nb2O5@N–C is less than that of T-Nb2O5, suggesting the low polarization of T-Nb2O5@N–C, which is propitious for practical demand. Compared with T-Nb2O5, the carbon-coating sample shows better rate performances, as shown in Fig. 4c. The discharge capacitances of T-Nb2O5 decrease (185.1, 160.6, 150.8, 132.3 and 110.2 mAh·g−1) with current density increasing from 1C to 20C. At the rates of 1C, 2C, 5C, 10C and 20C, T-Nb2O5@N–C shows reversible capacities of 221.7, 189.6, 184.3, 172.1, and 155.5 mAh·g−1, respectively. When current density returns to 1C, the electrode recovers 204.6 mAh·g−1, revealing the excellent reversible capability of T-Nb2O5@N–C. The cycling performance at high current density is an important parameter to further illustrate lifespan of the electrode in practical application. Impressively, T-Nb2O5@N–C shows excellent longevity life and superior capacity retention (Fig. 4d). The initial cycle of T-Nb2O5@N–C capacitance is 176.4 mAh·g−1 at 10C, and it keeps a capacity of 101.1 mAh·g−1 at the 2000 cycles with a higher capacity retention of 57.3%. However, after the first cycle (147.6 mAh·g−1), the specific capacity of T-Nb2O5 continues to decrease until the 100th cycle is stable. The capacity of T-Nb2O5 only remains 42.6 mAh·g−1 after 1500 cycles with a capacity retention of 28.8%. In addition, the Coulombic efficiency of T-Nb2O5@N–C is found to be nearly 100% in each cycle. Table 1 shows the comparison of electrochemical performances of some Nb2O5-based anode materials, which are taken from the previous typical reports with excellent performances in LIBs [18, 19, 30, 37, 38]. This result demonstrates that the active site of N atom and conductive carbon material can not only improve the effective electron transmission path for the high rate capability, but also inhibit structure pulverization after long-periodic cycling.

Fig. 4
figure 4

Galvanostatic discharge/charge curves of a T-Nb2O5 and b T-Nb2O5@N–C at 1C, 2C, 5C, 10C, 20C current densities; c rate performance and d cycle performance of T-Nb2O5 and T-Nb2O5@N–C; e EIS spectra of T-Nb2O5 and T-Nb2O5@N–C and corresponding equivalent circuit

Full size image
Table 1 Comparison of electrochemical performance for the same reported Nb2O5-based materials
Full size table

EIS and CV measurements of T-Nb2O5@N–C were further investigated to analyze electrode kinetics for Li-storage. According to the equivalent circuit fitting (Fig. 4e), the calculated charge transfer resistance (Rct = 102 Ω) is smaller than that of T-Nb2O5 (Rct = 204 Ω). It is suggesting that the enhanced electric conductivity of T-Nb2O5@N–C is attributed to the surface modification of N-doped carbon. The effect of N-doped carbon network on the electrical properties of T-Nb2O5@N–C can be researched by the four-point probe setup. The electrical conductivity of the T-Nb2O5@N–C can be one or two orders of magnitude better than that of T-Nb2O5 at the same intensity of pressure (Fig. S2), due to the addition of N-doped carbon network. Therefore, the increased electrical conductivity in this work clearly supports the formation of additional conductive paths between N-doped carbon network and T-Nb2O5. CV curves of T-Nb2O5 and T-Nb2O5@N–C with a voltage range of 1.0–3.0 V are shown in Fig. 5. Clearly, two oxidation peaks at 1.85 and 1.98 V and two strong reduction peaks at 1.85 and 1.53 V of the two samples both can be observed in Fig. 5a at 0.1 mV·s−1 [39]. The main electrochemical reaction is between Nb5+ and Nb3+ in the process of intercalation/deintercalation of Li+. The two reduction peaks are corresponding to the valence state from Nb5+ to Nb3+, indicating the Li+ intercalation into the Nb2O5. On the contrary, the oxidation peaks reflect lithium-ion extraction process of LixNb2O5. The specific process is shown in Eq. (1):

$${\text{Nb}}_{{2}} {\text{O}}_{{5}} + x{\text{Li}}^{ + } + x{\text{e}}^{ - } \Leftrightarrow {\text{Li}}_x{\text{Nb}}_{2} {\text{O}}_{5},$$
(1)

where x ≤ 2 (corresponding to 200 mAh·g−1). The main reduction and oxidation peaks location could be explained to the out and in of Li+ from orthorhombic Nb2O5 crystal structure.

Fig. 5
figure 5

a CV curves of T-Nb2O5 and T-Nb2O5@N–C at 0.1 mV·s−1; CV curves of b T-Nb2O5 and c T-Nb2O5@N–C at different scan rates from 0.2 to 5.0 mV·s−1; d b values determined by the relationship between peak current and sweep rate of T-Nb2O5@N–C; e capacitive contribution (green shaded regions) in CV test of T-Nb2O5@N–C at 0.2 mV·s−1; f contributions of capacitive and diffusion-controlled capacities at different sweep rates of T-Nb2O5@N–C

Full size image

To further explore the kinetics of T-Nb2O5@N–C electrode, the CV curves at different scan rates were carried out on T-Nb2O5 and T-Nb2O5@N–C (Fig. 5b, c). At the same sweep speed, CV curves of the two samples have similar peak patterns, suggesting the same obvious redox reactions along with Li+ insertion/extraction. It is worth noting that T-Nb2O5@N–C electrode exhibits much smaller polarization than T-Nb2O5 at 5.0 mV·s−1, demonstrating better electrochemical reversibility for T-Nb2O5@N–C at high current density. As reported by previous studies for CVs [37, 40], the total charge storage of electrodes could be separated into two parts: the contribution of Li+ insertion/extraction controlled by diffusion and the surface-controlled capacitive contribution. The linear relationship between the peak currents (i) and the scan rates (v) in log operation follows the equations:

$$i = av^{b}$$
(2)
$$\lg i = \lg a + b\lg v$$
(3)

where b value can provide the charge storage mechanism during the electrochemical process. The b value approaching to 0.5 means that the process is completely limited by diffusion. Contrarily, the capacity is controlled by surface-controlled capacitive behavior when the b value accesses to 1. As shown in Fig. 5d, the slope b values of cathode and anode are both 0.92, illustrating that the electrochemical reactions of T-Nb2O5@N–C are mainly dominated by the surface-controlled capacitive behavior (pseudocapacitive). Furthermore, Eq. (4) was adopted to quantitatively analyze the contributions of diffusion and capacitive behavior [40].

$$i = k_{1} v + k_{2} v^{1/2}$$
(4)

where k1v and k2v1/2 are represent the surface-controlled capacitive behavior and diffusion-dominated process, respectively. Therefore, the pseudocapacitive contribution ratio can be calculated by k1v/i. According to the analysis, pseudocapacitive contribution raises 56.4% of the capacity storage of T-Nb2O5@N–C at 0.2 mV·s−1, as demonstrated in Fig. 5e (green region). Moreover, the pseudocapacitive contribution gradually further enlarges from 56.4% to 91.1% with the scan rate increasing from 0.2 to 5.0 mV·s−1, as shown in Fig. 5f. In contrast, the capacitive contributions are extremely higher than those of T-Nb2O5 (Fig. S3). More importantly, T-Nb2O5@N–C exhibits pseudocapacitance dominant charge transfer behavior, resulting in the long-life cyclability and superior rate performance.

For the actual application of the as-prepared material, full cell is necessary to be investigated. LiNi0.5Mn0.3Co0.2O4 was designed as the cathode to match with T-Nb2O5@N–C anode. The LiNi0.5Mn0.3Co0.2O4/Li half cells show a reversible capacity of 172.5 mAh·g−1 at 0.2C within 2.75–4.20 V (Fig. 6a). The mass of cathode material is excessive, so the performance of T-Nb2O5@N–C material is the key to the capacity in the full cell. Figure 6b, c shows large reversible capacities of 175.7, 175.5, 167.9, 159.2, 144.5 and 123.6 mAh·g−1 at 0.2C, 0.5C, 1.0C, 2.0C, 5.0C and 10.0C in full cell, respectively. After 100 cycles, up to 95.0% of the initial capacity is retained (Fig. 6d). When cycled at such a high current density (5C), the full cell still achieves higher capacity retention of 77.7% over 500 cycles, as shown in Fig. 6e. The rate capacity of this work is more superior to those previous reports [41, 42]. Therefore, T-Nb2O5@N–C is expected to become a practical anode material for the LIBs of electric vehicles.

Fig. 6
figure 6

a Charge–discharge curves of LiNi0.5Mn0.3Co0.2O4/Li and T-Nb2O5@N–C/Li half cells at 0.2C; b charge–discharge curves and c rate performance of LiNi0.5Mn0.3Co0.2O4||T-Nb2O5@N–C full cell at different current rates; cyclability of LiNi0.5Mn0.3Co0.2O4||T-Nb2O5@N–C full cell at d 1.0C over 100 cycles and e 10.0C over 500 cycles

Full size image

4 Conclusion

In summary, the porous ultrathin N-doped carbon-coated T-Nb2O5@N–C was fabricated by a sample and effective approach with ionic liquids as a carbon and nitrogen source. T-Nb2O5@N–C presents excellent rate performance with a discharge capacity of 155.5 mAh·g−1 at 20C. Meanwhile, it exhibits a long-term cycling life with a discharge capacity of 101.1 mAh·g−1 at 10.0C after 2000 cycles. It is confirmed that ultrathin N-doped carbon coating for T-Nb2O5 not only endows the fast electrons and ions transport channels and buffers the volume changes under the long-term cycling process, but also can further yield excellent Li+ intercalation pseudocapacitive property with high rate capability. The LiNi0.5Mn0.3Co0.2O4||T-Nb2O5@N–C full cell also exhibits good electrochemical performance. Therefore, the proposed synthesis methodology gives a simple, effective and scalable way to construct carbon-based composites for practical application.